8 
all areas of the spatial assessment unit; providing equal weight to such data could 
bias the assessments. 
A second advantage is that the CFD incorporates the spatial-temporal pattern of 
criteria exceedance into the assessment. The shape of the curve offers information on 
patterns of exceedance in space and time. Such information may prove helpful in 
understanding the causes of impairments (see page 162 in U.S. EPA 2003a). 
A third advantage is that it bases the assessment on biologically determined patterns 
of allowable criteria exceedance. Reference curves are ideally developed in the same 
way as assessment curves and should reflect the degree of criteria exceedance that 
can be withstood by the ecological communities without impairing the designated 
use. Thus, comparison of the assessment curve to the reference curve ensures that 
any allowable criteria exceedances do not occur in a spatial or temporal pattern that 
could, in reality, represent impairment at the scale of the entire assessment unit (see 
pages 162-178 in U.S. EPA 2003a). Local persistent effects could still have high 
impairment. 
Finally, the combined elements of the CFD criteria assessment methodology fully 
and effectively address all five factors used to determine attainment of designated 
uses: magnitude, duration, frequency, space, and time. After conducting a national 
review of TMDL programs, the National Research Council (2001) concluded that 
establishing these conditions is crucial for successful application of state water 
quality standards. 
The CFD methodology is a new and innovative method of water quality criteria 
assessment, representing an improvement over methods used in other parts of the 
country (STAC 2006). The standard practice for assessing compliance with water 
quality criteria throughout the United States is by sampling monthly at a fixed set of 
stations and gauging compliance strictly from a count of exceedances of those 
samples. Sampling stations are typically located for convenience (e.g., accessibility). 
Consequently, reluctance to re-evaluate and change location (so as to maintain a time 
series at a fixed point) is common; no consideration is given to the representative¬ 
ness of the sample for the space/time not sampled. 
Most assessments are based simply on EPA assessment guidance in which all 
samples in a given area were compiled; attainment was assumed if no more than 10 
percent of the samples exceeded the standard (U.S. EPA 1997). In this approach, all 
samples are assumed to be fully representative of the specified space and time and 
are simply combined as if they were random samples from a uniform population. 
This approach was necessary in the past because the technology did not exist for a 
more rigorous method of data analysis; however, it neglected spatial and temporal 
patterns in the criteria measures. The CFD approach was designed to characterize 
these spatial and temporal patterns and weight samples more accurately based on the 
amount of space or time that they actually represent. 
chapter i 
Introduction 
